/ propagandalytics

Just how far did Russian propaganda go?

Facebook and Twitter have changed their tune on Russian propaganda again. In prepared remarks before congress tomorrow, Twitter will state that they found evidence that 36,746 accounts "generated approximately 1.4 million automated, election-related Tweets, which collectively received approximately 288 million impressions" in the run up to the 2016 US election. Likewise, Facebook will admit that election-oriented Russian disinformation reached approximately 126 million users in 2016.

But is this the full extent of it?

It seems that every week or two there is a new admission that foreign influence operations around the 2016 US election were more pervasive than before. And despite Facebook and Twitter revealing more data as time goes on, they are not publishing research. They have claimed that journalism and academic research into social-media propaganda is fatally flawed ― because of the limits that the platforms themselves have built into the systems available to researchers. But Twitter and Facebook have not published rigorous research that supports their claims. Instead, they have periodically offered nuggets of data in attempts to do damage control. Or, in propaganda-researcher terminology, to minimize the import and attention given to the story. And according to Zeynep Tufekci, that purposeful minimization of attention is the twenty-first century's form of censorship. Denying attention is how you silence an unwanted narrative in the age of information abundance.

So what is the truth in all of this?

From my perspective, there are two important takeaways from all of this.

First, this is not the last admission of foreign social-media influence operations during the 2016 campaign. Neither the data I've examined, the research other scholars have published, nor the platforms' admissions suggest that this is the end of the story. And until things change on the platforms, these influence operations will continue to take place, and to succeed.

Second, and more importantly, the numbers shared by Twitter and Facebook are disingenuously low. In both cases, they provide the number of people who saw (or at least scrolled by) posts from known bad actors. But that's the minimum threshold of exposure. What's important is not the specific post but the narrative.

When I see a social media post, it has a psychological impact. Familiarity breeds truthiness ― especially when that familiarity is grown unconsciously, like it is when we casually scroll through our social feeds. It makes other posts similar to it, but from other sources, more believable. And it makes us more likely to generate our own posts (and offline claims) that are similar to it. There's a reason we call these influence operations ― not engagement operations or impressions operations. It's not about the reach of the specific post ― that's what people looking to make money off of ads care about. It's about the reach of the narrative. And if that narrative can be laundered through other social-media platforms and users so that people don't know the real source of the narrative, that's even better!

As my colleagues and I have argued before, it's control of the narrative and disruption of the democratic process that these propagandists are after, not convincing people that a story is true. Sow the seeds of doubt in the political candidates, the political process, the government, the media, academia, ... and you've gaslit the country.

To my mind, the end goal of Russian election propaganda was probably not to get Donald Trump elected. The goal was to get us having the very debate we're having right now, in which I am participating by writing this post. To get our congressional intelligence committees interrogating American business executives instead of investigating foreign agents. To eliminate the benefits of the first amendment by amplifying the most poisonous of free speech online and disempowering the press.

I'm glad that this information is coming out, that congress is investigating, that they have good data people working for them, that they're holding social media executives' feet to the fire, that they're calling foul when they make erroneous claims. But this isn't about posts, accounts, impressions, or engagements. It's about influence ― which is orders of magnitude larger than even the most liberal estimates of impressions. It's about the credibility of our government, our press, and the fundamental freedoms guaranteed in our constitution.

Our individual and collective psyches are under attack. Our constitution is under attack. Our democracy is under attack. And I sincerely hope that the outcome of these congressional hearings is that Washington, Silicon Valley, and Academia will stop fighting each other, and start partnering together in the fight to preserve the Republic and the legitimacy of the ideas on which it ws founded.